Level of Perinatal Care of the Maternity Unit and Rate of Cesarean in Low-Risk Nulliparas.

Abstract


OBJECTIVE:

To analyze the influence of level of perinatal care of the maternity unit on the rate of cesarean delivery during labor among women with low-risk pregnancies.

RESULTS:

Overall, the rate of cesarean during labor was 11.7%. The size and status of the facilities did not significantly affect these rates. Risk factors for cesarean were older maternal age, non-French origin, gestational age of 41 weeks, male sex, and high birth weight.

CONCLUSION:

Maternity units that frequently manage high-risk pregnancies have higher rates of cesareans during labor for their population of nulliparas at low risk than do facilities that deal mainly with low-risk pregnancies

Introduction

Management of pregnant women at low risk and in particular their cesarean rate is a topic of concern today. Between 1998 and 2003 the cesarean rate increased from 17.5% to 20.2% in France. As the authors of the French Perinatal Assessment (an expert white paper commissioned by the government to recoMme.nd improvements in perinatal care) pointed out that “While the need for intensive care is obvious in high-risk situations, the question is far more open when the risk is low. In such situations, excess surveillance is sometimes iatrogenic. The available data suggest that we simultaneously need to do more and better in high-risk situations and less (and better) when the risk is low.”

In 1998, the government enacted regulations requiring the classification of maternity units according to the level of perinatal care available there. This classification was based on neonatal care. In France, however, a high level of neonatal care is most often correlated with a high level of maternal care, because management of high-risk pregnancies frequently entails management of high-risk neonates. This classification therefore reflects the level of perinatal care at the maternity unit. Women at low risk in France can choose to deliver in any hospital, whatever its level of care. Level 2b and 3 units are intended to provide management for high-risk pregnancies, but remain responsible for providing care for the patients in their geographic sector, including pregnancies at low risk. Moreover, some women without any risk factors want to ensure the “maximum safety” for their neonates and choose to give birth in a maternity ward with a neonatal critical or special care unit. The question thus arises: does management of pregnancies at high risk, associated with a high but justifiable cesarean rate, affect the same team’s management of women at low risk?

Groups in different countries have long examined the association between the status (private or public) or size of maternity units and their cesarean rates. The level of perinatal care associated with the maternity ward (levels 1, 2a, 2b or 3) is a relatively recent concept, but it may be more relevant in analyzing the influence of the management of high-risk pregnancies on the cesarean rates among low-risk patients.

We used data from the PREMODA (PREsentation et MODe d’Accouchement: presentation and mode of delivery) study, which involved 138 French maternity units, to estimate whether variations in the rates of cesareans during labor among nulliparas at low risk were associated with characteristics of the place of delivery and, in particular, with the level of perinatal care available there.

DISCUSSION

In our study, the rate of cesareans during labor among nulliparas at low risk varied according to the level of perinatal care available at the maternity unit. Hospitals that frequently manage high-risk pregnancies (level 2b and 3) performed cesareans during labor for this population at a higher rate than those that deal mainly with low-risk pregnancies (level 1). The level of the maternity unit was the only structural factor identified in our study as a significant risk factor for cesarean during labor. The level of perinatal care available at the maternity unit is relevant to the quality of care and is therefore information that should be collected for studies seeking to compare the quality of care at different units.

Criteria for defining low-risk nulliparas are not homogeneous from study to study. Some assess the risk at the beginning of pregnancy, at the first obstetric consultation. But authors most often use the definition by Cleary et al of the standard nullipara at the end of pregnancy, a definition similar to ours. We excluded cesareans before onset of labor because as a general rule these are performed for nulliparas in France only in pregnancies with complications. “Convenience c-sections” remain a marginal practice in France. In PREMODA, of 4,218 nulliparas with a cephalic fetus at term, only 101 (2.4%) had a cesarean before labor, and 98 were performed for medical indications. The very low percentage of maternal and neonatal complications in our population of low-risk nulliparas confirms the validity of the inclusion and exclusion criteria chosen to define the nulliparas at low risk. Moreover, our cesarean rate is close to those of other recent studies in low-risk nulliparous populations.

The sample of the PREMODA control group is representative within each maternity ward because it selected every singleton in cephalic presentation delivered at term with a delivery log number that was a multiple of 20. This study was based on prospective data collected rigorously; local investigators from every maternity unit were responsible for checking its exhaustiveness and quality. They sent the data regularly to a regional coordination office, which sent it on to the national office responsible for its prospective analysis. At the end of the study, a national coordinator (Marion Carayol) visited 22 facilities, each selected randomly within a region, to assess the quality of data collection.

Our data include deliveries from 138 establishments representing more than one fifth of all French maternity units. Although our study is large, it includes only a sample of French maternity units, and thus, our results cannot necessarily be extrapolated to all French maternity units: the geographic distribution and structural characteristics of the units in our study differ from those of all maternity units in France. The 2003 national perinatal survey found that 36.5% of births took place in level 1 and 18.8% in level 3 maternity units (compared with 18.4% and 36.6%, respectively, in our study). Comparisons with this national survey are difficult, however, because its results concern all women giving birth and not the subgroup of low-risk nulliparas. Another French multicenter study found that 31% of nulliparas at low risk delivered in level 3 units.

The PREMODA study is based on the willingness of maternity units to volunteer to provide data. The facilities that participate in our study were therefore probably more motivated to assess their obstetric practices than French maternity units as a whole. This is particularly true for the participating level 1 units. Although level 2 and 3 facilities are used to employing and evaluating protocols for the management of problem pregnancies, level 1 units face such assessment problems less often. We therefore think it likely that the level 1 units that agreed to participate in our study are not totally representative of all level 1 facilities. Nevertheless, the number and diversity (level, status, and size) of the participating units and the quality of the data collection allow us to think that our results reflect the general practices of the obstetric teams in different facilities. Although the study population is not strictly representative of the French population, it provides us with sufficient power to analyze the relation between level of perinatal care and the rate of cesareans during labor.

Several hypotheses may explain the differences in these cesarean rates as a function of level of perinatal care. Perhaps, in high-risk facilities, physicians are more likely to expect problems and may encourage the use of cesarean deliveries. On the other hand, in low-risk facilities, midwives expect successful vaginal delivery and use active management of labor to achieve it, as seen by the high rate of oxytocin use. Despite their higher induction rates, level 1 maternity units had a significantly lower rate of cesareans during labor than the higher-rated facilities. Moreover, management of medical interventions during labor seemed to differ according to the level of care: obstetric teams in level 1 units used oxytocin significantly more often than those in other maternity units. It is possible that labor proceeds better in level 1 units. Meta-analyses have shown that “active management” of labor in nulliparas can reduce its duration and tends to reduce the percentage of cesarean deliveries, although not quite significantly.

On the other hand, midwives in level 1 units manage labor only for low-risk women and may be more available to these women than are the midwives in level 2 and 3 units, who are responsible simultaneously for patients at higher risk, including in utero transfers and premature deliveries. Support for women during labor is essential for the success of vaginal delivery. Numerous studies show that midwives’ continuous presence has a positive effect on both labor (diminution of its duration and of analgesia use) and delivery(reduction in instrumental interventions and caesareans). We thus suggest 2 hypotheses for the success of vaginal delivery in level 1 facilities: the effective management of interventions during labor and the availability of midwives. The results we observed, that is, more interventionism but fewer cesareans during labor, are not really paradoxical in the French context, where obstetricians’ attention to medicalization seems to focus primarily on decreasing the cesarean rate, rather than on any of its other aspects.

Another hypothesis centers on the availability of certain types of personnel or on the organization of care in maternity units by level of care. If it is easier to perform a cesarean delivery a in level 3 or 2b unit because of the availability of personnel (such as anesthesiologists or nursing support staff), this might affect obstetric decision-making. This has not been studied in relation to cesareans, but the organization of care has been found to influence medical practices in other situations, such as the transfer of newborns to neonatal units.

The patient’s choice of maternity ward can also in itself serve as an explanation. Some low-risk patients prefer to give birth in a maternity ward with a NICU or special care nursery to maximize their newborns’ safety. These patients, especially well informed, may be more likely to have a cesarean during labor than the less anxious patients willing to give birth in level 1 facilities. The mother’s socioeconomic situation does not seem in other studies to be associated with the rate of cesareans during labor. A recent study of the criteria for the choice of delivery facility among low-risk patients in France showed that this decision has multiple determinants that vary according to region, maternal age, parity, educational level, and national origin. Some of these factors in turn are associated with the rate of cesareans during labor. Geographic origin appears to play a role in the risk of cesareans during labor, which occur at a higher rate among women born outside of France. Their lower socioeconomic status, perhaps mediated by less good prenatal care, may explain this finding. Others have reported a similar relation with the mother’s country of birth in populations of low-risk nulliparas. The other risk factors for cesarean delivery during labor (maternal age, term gestation, birth weight, and male sex) were the same as those usually reported in the literature.

Cesarean deliveries in nulliparas are the principal factor in the global augmentation of cesareans observed since the end of the 1970s. Better management of nulliparas at term should make it possible to reduce the primary cesarean rate and therefore the global rate, as some authors have shown. If this is our aim, then level 2b and 3 maternity units must assess their management of nulliparas at low risk to attempt to explain the differences observed in comparison with level 1 units. Certainly, we cannot assess the quality of a maternity unit solely by its caesarean rate, and an overall assessment would require us to look as well for the possible harmful effects of a low cesarean rate on mothers and babies. Our study lacks the power to answer this question, given the low frequency of complications (0.1% of maternal intensive care transfers and 1.0% neonatal). Nonetheless, the level 1 maternity units must also look into the possible consequences, in particular neonatal, of their low cesarean rates. The price of reducing the cesarean rate in low-risk nulliparas must not be an increase in the neonatal risk.

The comparison of practices between maternity units is a starting point for an assessment in each unit aimed at improving the management of low-risk nulliparas during labor and thereby, perhaps reducing the cesarean rate. Currently, there is no consensus about the management of nulliparas at low risk and in particular about the “optimal” cesarean rate. The organization of care of high-risk pregnancies may have an impact on the care of women with uncomplicated pregnancies. A reflection about the place of low-risk women in the level 2b and 3 maternity units should be engaged to improve their management and optimize their cesarean rate.